But you don't know how many db spl you have as an output from this.
Your mic/preamp* ( if you use a 'regular' kind of setup*) doesn't know where the mic is located nor if the signal captured by the mic is at a defined level.
If you don't have a known reference ( acoustic) your software can't know if the signal it acquire is either 105dbspl, 60 or 90.
* i don't talk about your 'line' level hifi preamp but the preamplifier following your mic in the record chain.
Google Image Result
Forget eq/compressor/and vocals 😉
Your mic/preamp* ( if you use a 'regular' kind of setup*) doesn't know where the mic is located nor if the signal captured by the mic is at a defined level.
If you don't have a known reference ( acoustic) your software can't know if the signal it acquire is either 105dbspl, 60 or 90.
* i don't talk about your 'line' level hifi preamp but the preamplifier following your mic in the record chain.
Google Image Result
Forget eq/compressor/and vocals 😉
Last edited:
The REW and UMIK-1 is a calibrated system. The mic is a digital interface to the software, the mic's adc is factory calibrated and each mic has a calibration file.
Naturally I have to trust that or benchmark that against a calibrated reference, but for now I have to trust the factory.
Edit: I do have another SPL meter I will try.
Naturally I have to trust that or benchmark that against a calibrated reference, but for now I have to trust the factory.
Edit: I do have another SPL meter I will try.
Last edited:
Ok this is what i thougths from our last exchange.
It is calibrated yes but at which distance from the source, do you use the same condition as in the factory? Same principle apply about loss of level and distance than with your loudspeaker with your mic.
Your software needs a reference voltage to apply it's own range on the measurement you perform.
These 'integrated' solutions often leads to 'wrong' readings because you don't have (easily) access to all parameters at play. Too much simplification.
One thing you can do to be in the ballpark is to use either a sonometer application on your phone or a cheap SPL meter* to be sure your measurements are in the range more or less.
*Something like that should be ok:
Mengshen Sonometre 30-130dBA, Decibelmetre Professionnel avec ecran Retro-eclaire (Pile De 9V Incluse): Amazon.fr: Bricolage
Edit: ok if you have another spl meter then use pinknoise rather than a sinus at 1k. It should be easier for these small sonometer to analyse. Then try to match level between both your software and the splmeter it'll be enough to have a close to reality reading.
It is calibrated yes but at which distance from the source, do you use the same condition as in the factory? Same principle apply about loss of level and distance than with your loudspeaker with your mic.
Your software needs a reference voltage to apply it's own range on the measurement you perform.
These 'integrated' solutions often leads to 'wrong' readings because you don't have (easily) access to all parameters at play. Too much simplification.
One thing you can do to be in the ballpark is to use either a sonometer application on your phone or a cheap SPL meter* to be sure your measurements are in the range more or less.
*Something like that should be ok:
Mengshen Sonometre 30-130dBA, Decibelmetre Professionnel avec ecran Retro-eclaire (Pile De 9V Incluse): Amazon.fr: Bricolage
Edit: ok if you have another spl meter then use pinknoise rather than a sinus at 1k. It should be easier for these small sonometer to analyse. Then try to match level between both your software and the splmeter it'll be enough to have a close to reality reading.
Last edited:
Did another measurement tonight. Using the UMIK-1 and an old Radioshack SPL meter I carefully put both devices on axis as close to 1m as possible. Since the speaker uses a stepped baffle I used 1m from the midrange driver using 1 khz tone. (I will try pinknoise later).
The scope I am using has a true RMS measure function so I turned up the sound until I got as close to 2.82Vrms as I could. The UMIK-1 and the RS SPL meter were within 2db of each other at 82-84db.
The scope I am using has a true RMS measure function so I turned up the sound until I got as close to 2.82Vrms as I could. The UMIK-1 and the RS SPL meter were within 2db of each other at 82-84db.
Yes, the UMIK-1 has sensitivity in the calibration file.
Assuming you've got 2.83 volts at 1 meter (from the midrange's dust cap) then that's your sensitivity.
It's not surprising that it's such a low sensitivity because the midrange is a real midrange (not a midbass driver) and likely isn't being used below about 300 Hz. This leaves the midbass driver to do all the "lifting" with respect to baffle-step loss compensation. So if the midbass is around 88 db you are likely down another 5-6 db from that to compensate for baffle-step loss.
Look to the Elsinore loudspeaker as a potential replacement: it should get you near your sensitivity goal and has a very easy impedance. Remember that SE amps are reactive to impedance, so just "any" loudspeaker that's "voltage-linear" ("flat" freq. response with low out impedance amplifier) will likely have freq. linearity issues with an SE amp (as the amp reacts to "swings" in impedance with freq.: changing the freq. response). In other words: just having a higher efficiency loudspeaker isn't enough (..unless you are using digital eq. to compensate for freq. response changes).
Assuming you've got 2.83 volts at 1 meter (from the midrange's dust cap) then that's your sensitivity.
It's not surprising that it's such a low sensitivity because the midrange is a real midrange (not a midbass driver) and likely isn't being used below about 300 Hz. This leaves the midbass driver to do all the "lifting" with respect to baffle-step loss compensation. So if the midbass is around 88 db you are likely down another 5-6 db from that to compensate for baffle-step loss.
Look to the Elsinore loudspeaker as a potential replacement: it should get you near your sensitivity goal and has a very easy impedance. Remember that SE amps are reactive to impedance, so just "any" loudspeaker that's "voltage-linear" ("flat" freq. response with low out impedance amplifier) will likely have freq. linearity issues with an SE amp (as the amp reacts to "swings" in impedance with freq.: changing the freq. response). In other words: just having a higher efficiency loudspeaker isn't enough (..unless you are using digital eq. to compensate for freq. response changes).
Last edited:
Oops. After some hours of sleeps it occured to me i was plain wrong and i blamed your mic instead of my tired brain jderimig!
Please apologize and i hope i didn't induce confusion on your side. 🙄
ScottG answer about bsc makes much more sense.
Hi,
82dB/2,8V/1m is not that uncommon without room gain, although I would have estimated your speakers a little bit higher. But the real trouble for your SET is probably, as Scott said, the impedance variation caused by the multiway crossover.
Real 92dB/2,8V/1m (many manufacturers are very optimistic about their efficiency numbers, and/or assume a certain room gain to blow them up) is quite hard to achieve with a decent sized box, and will be limited in bass range. If you like this number also with deep bass, your efforts will raise and the box will be very large, with large bass drivers.
This would mean a multi-way box again, and you´ll need to take care about the impedance variations. It´s doable for sure... but depending on your volume needs, room size, speaker placement and taste a little bit more simple solution like a decent FR driver might be an option. Doing that right does not mean to simply screw down a driver in a box and be done... and even if done right, might end up a few dB below 92. And will not do deep bass, and will not play extremely loud...
All the best
Mattes
82dB/2,8V/1m is not that uncommon without room gain, although I would have estimated your speakers a little bit higher. But the real trouble for your SET is probably, as Scott said, the impedance variation caused by the multiway crossover.
Real 92dB/2,8V/1m (many manufacturers are very optimistic about their efficiency numbers, and/or assume a certain room gain to blow them up) is quite hard to achieve with a decent sized box, and will be limited in bass range. If you like this number also with deep bass, your efforts will raise and the box will be very large, with large bass drivers.
This would mean a multi-way box again, and you´ll need to take care about the impedance variations. It´s doable for sure... but depending on your volume needs, room size, speaker placement and taste a little bit more simple solution like a decent FR driver might be an option. Doing that right does not mean to simply screw down a driver in a box and be done... and even if done right, might end up a few dB below 92. And will not do deep bass, and will not play extremely loud...
All the best
Mattes
Last edited by a moderator:
I'd suggest you make a normal sweep measurement with REW, and look at the resultant freq response curve. Because sensitivity at 1kHz may or may not represent a good nominal value.
Set the output level in REW to have the amp make 2.83v across the sweep.
I get anywhere close to 2.83v, and just do the math to adjust to 2.83v/1m
Decibels to Voltage Gain and Loss convert calculation conversion amplification amplifier electronics - sengpielaudio Sengpiel Berlin
Sound level distance damping decibel dB damping calculation calculator change distance versus sound level apps reduction drop dissipation SPL sound transmission loss free field loss sound and distance - decrease drop fall sound over distance versus d
Better yet, and best method I've found for measuring sensitivity, is run full bandwidth pink noise through the speaker, measure the voltage...and again do the math.
Pink voltage bounces around a little, so use time averages of both voltage and SPL, taken over the same time interval. (30 sec to 1 minute works fine).
REW's SPL meter set to Leq will give the SPL average, your scope should probably handle the voltage if it's digital.
(I typically use a true RMS multimeter.)
This is a no BS measure of sensitivity imo, as it's essentially an integrally weighted measurement.
Measure current over the same time interval, and you can derive a true nominal impedance.
Set the output level in REW to have the amp make 2.83v across the sweep.
I get anywhere close to 2.83v, and just do the math to adjust to 2.83v/1m
Decibels to Voltage Gain and Loss convert calculation conversion amplification amplifier electronics - sengpielaudio Sengpiel Berlin
Sound level distance damping decibel dB damping calculation calculator change distance versus sound level apps reduction drop dissipation SPL sound transmission loss free field loss sound and distance - decrease drop fall sound over distance versus d
Better yet, and best method I've found for measuring sensitivity, is run full bandwidth pink noise through the speaker, measure the voltage...and again do the math.
Pink voltage bounces around a little, so use time averages of both voltage and SPL, taken over the same time interval. (30 sec to 1 minute works fine).
REW's SPL meter set to Leq will give the SPL average, your scope should probably handle the voltage if it's digital.
(I typically use a true RMS multimeter.)
This is a no BS measure of sensitivity imo, as it's essentially an integrally weighted measurement.
Measure current over the same time interval, and you can derive a true nominal impedance.
Excellent suggestion!
I have done that before when trying to equalize for my room. I know there is about a 3db dip at 1 Khz.
The only concern about pink noise is whether I will measure a little low frequency component room gain in the results. I'd do that outside if I were younger.
Edit: I have also adjusted the tweeter and midrange pots on the speaker to give a Harman curve in my room at my seating position awhile ago. I do not think this affected the 1khz point though.
I have done that before when trying to equalize for my room. I know there is about a 3db dip at 1 Khz.
The only concern about pink noise is whether I will measure a little low frequency component room gain in the results. I'd do that outside if I were younger.
Edit: I have also adjusted the tweeter and midrange pots on the speaker to give a Harman curve in my room at my seating position awhile ago. I do not think this affected the 1khz point though.
Last edited:
Yes, the indoor room gain is an issue.
I do go outside for this, well that is for as long as my 69yr old butt can still lug stuff outside !
I do go outside for this, well that is for as long as my 69yr old butt can still lug stuff outside !
if you love audio you'll be lugging stuff till the cows come home or your pushing the daisies from the other side...
Yup, 20 years ago I would be dragging these 100# speakers outside and P.O.ing off the neighbors.
well likely no tours again this season so moving my home gear is what's keeping me in shape...
100 pounds that's peanuts compared to a Clair S4.
100 pounds that's peanuts compared to a Clair S4.
Ok this time i had sleep last night and i'm sure about would i will state ( 😉 ) :
i emphized about the use of a sonometer because cheap ones are (for the one i've seen/used) usually DBa or give you the choice between A or C.
DbA filterout lowend (starting around 1khz) so they don't take roomgain into account in reading.
Dbc start filtering around 30hz.
Turk,
you move them alone?! I don't believe you!
i emphized about the use of a sonometer because cheap ones are (for the one i've seen/used) usually DBa or give you the choice between A or C.
DbA filterout lowend (starting around 1khz) so they don't take roomgain into account in reading.
Dbc start filtering around 30hz.
Turk,
you move them alone?! I don't believe you!
Last edited:
I still move these 56" tall, 280lb beasties alone.
....and even get 'em on top of 45" tall subs.
(with the mandatory help of a hydraulic lift table !!)
....and even get 'em on top of 45" tall subs.
(with the mandatory help of a hydraulic lift table !!)